Owners equivalent rent is the price the Bureau of Labor Statistics estimates a homeowner would pay to rent their own house. Nobody actually pays it. BLS constructs it from a rotating survey of renters and applies the result to the owner-occupied share of the housing stock. The number matters because owner-occupied housing is roughly two thirds of US households, and the CPI needs a price for the housing service those households consume even though no rent check changes hands.

The choice to impute a rent rather than track home prices or mortgage payments is deliberate. A house is a consumption good and an investment. The consumption piece is shelter services, which is what CPI is supposed to measure. The investment piece is capital appreciation, which is not consumption and does not belong in a consumer price index. BLS separates the two by pricing the shelter service the same way it prices a renter’s shelter service, using the rental market as the reference.

How BLS actually measures it

The Housing Survey is the input. BLS interviews about 50,000 rental units across the country on a six-month rotating panel. Each unit is visited twice a year, six months apart. The rent change between those two visits is the raw data.

Owners equivalent rent uses the subset of rentals that share structural characteristics with owner-occupied homes: single-family detached, two or three bedrooms, similar age and size. Tenant rent uses the full rental sample. Both series come out of the same survey, they just weight the sample differently.

Two features of the survey shape everything downstream. First, the six-month gap between visits means each observation is a six-month change, and BLS annualizes it. Second, most rents in the survey do not change at any given six-month observation because most leases run twelve months. New leases price to market. Existing leases do not until they renew.

The lag, and why it is roughly twelve months

Market rents move quickly. Zillow Observed Rent, Apartment List, RealPage, and the CoreLogic Single-Family Rent Index all track new leases signed in the current month and move within weeks of a demand or supply shift.

CPI shelter tracks the full stock of leases, most of which were signed six to eighteen months earlier. When market rents accelerate, only the new-lease share of the CPI sample reflects it in month one. The rest catches up as leases renew, roughly one twelfth of the sample per month.

The mechanical result is a lag of about twelve months between market-rent series and CPI shelter, with the peak correlation at ten to fourteen months depending on the specification. The 2021 to 2022 market-rent surge did not fully show up in CPI OER until late 2022 and 2023. The 2023 to 2024 market-rent deceleration did not fully close through CPI OER until early 2026.

That lag is the mechanism, not a bug. It reflects real economic reality: a household on a lease signed in 2023 is paying 2023 rent, not 2026 rent. CPI is supposed to price what households actually pay.

Why the CPI to PCE wedge opens on OER

Core CPI weighted OER at 33.7 percent through 2025. Core PCE weighted OER at 13.5 percent. A one percentage point difference in the OER print carries roughly 20 basis points more into core CPI than into core PCE.

The weight difference is not a methodology fight. PCE captures all consumer spending, including things households do not pay for directly, like employer-paid health insurance and Medicare. That expands the denominator and dilutes shelter’s share. CPI captures out-of-pocket spending only. Shelter is a bigger fraction of that.

Same OER series, two different weights, one predictable wedge. From July 2024 through December 2025, core CPI year-over-year ran an average 35 basis points above core PCE year-over-year. Owners equivalent rent accounted for most of the gap.

What to watch, and in what order

Real-time rent series move first. Zillow, Apartment List, and CoreLogic SFRI print monthly and give the earliest read on where CPI shelter will be in twelve to fifteen months.

BLS New Tenant Rent Index prints quarterly. This is BLS’s own attempt to publish a market-lease series inside the CPI framework, and it aligns closely with private trackers. It runs six to nine months ahead of the CPI OER print because it strips out the renewal share.

CPI shelter and OER print monthly with the headline CPI release, three-tenths of a percentage point on the shelter line moves core CPI by roughly one tenth. When Zillow and the New Tenant Rent Index agree that market rents have decelerated, the CPI OER print will follow, on the twelve-month clock.

The lag is predictable, the direction is knowable, and the wedge closes on its own timetable. The Fed knows all of this. Read the shelter line as a lagging confirmation, not a signal.